68 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1956 



INPUT 



FILTER 



AUTOMATIC 



GAIN 



CONTROL 



REGENERATOR 



DETECTOR 



TIMING 



WAVE 



GENERATOR 



FILTER 



OUTPUT 



Fig. 1 — A typical regenerative repeater shown in block form. 



That the Bell System is interested in the long-distance transmission 

 of television and other broad-band signals is evident from the number 

 of miles of such broad-band circuits, both coaxial cable and microwave 

 radio, ^ now in service. These circuits provide high-grade transmission 

 because each repeater was designed to have a very fiat frequency charac- 

 teristic and linear phase over a considerable bandwidth. Furthermore, 

 these characteristics are very carefully maintained. For a binary pulse 

 system employing regeneration the requirements on flatness of band and 

 linearity of phase can be relaxed to a considerable degree. The compo- 

 nents for such a system should, therefore, be simpler and less expensive 

 to build and maintain. Reduced maintenance costs might well prove to 

 be the chief virtue of the binary system. 



Since the chief advantage of a binary system lies in the possibility of 

 regeneration it is obvious that a very important part of such a system is 

 the regenerative repeater employed. Fig. 1 shows in block form a typical 

 broad-band, microwave repeater. Here the input, which might come from 

 either a radio antenna or from a waveguide, is first passed through a 

 proper microwave filter then amplified, probably by a traveling-wave 

 amplifier. The amplified pulses of energy are regenerated, filtered, am- 

 plified and sent on to the next repeater. The experiment to be described 

 here deals primarily with the block labeled "Regenerator" on Fig. 1. 



In these first experiments one of our main objectives was to keep the 

 repeater as simple as possible. This suggests regeneration of pulses 

 directly at microwave frequency, which for this experiment was chosen 

 to be 4 kmc. It was suggested by J. R. Pierce and W. D. Lewis, both of 

 Bell Telephone Laboratories, that further simplification might be made 

 possible by accepting only partial instead of complete regeneration. 

 This suggestion was adopted. 



For the case of complete regeneration each incoming pulse inaugurates 

 a new pulse, perfect in shape and correctly timed to be sent on to the 



'A. A. Roetken, K. D. Smith and R. W. Friis, The TD-2 System, B. S. T. J., 

 Oct., 1951, Part II. 



